The invasion of novel habitats is recognized as a major promotor of adaptive trait evolution in animals. We tested whether similar ecological niches entail independent and adaptive evolution of key phenotypic structures related to larval host invasion in distantly related taxa. We use disparately related clades of coral barnacles as our model system (Acrothoracica: Berndtia and Thoracica: Pyrgomatidae). We analyze the larval antennular phenotypes and functional morphologies facilitating host invasion. Extensive video recordings show that coral host invasion is carried out exclusively by cypris larvae with spear-shaped antennules. These first exercise a series of complex probing behaviors followed by repeated antennular penetration of the soft host tissues, which subsequently facilitates permanent invasion. Phylogenetic mapping of larval form and function related to niche invasion in 99 species of barnacles (Thecostraca) compellingly shows that the spear-phenotype is uniquely associated with corals and penetrative behaviors. These features evolved independently in the two coral barnacle clades and from ancestors with fundamentally different antennular phenotypes. The larval host invasion system in coral barnacles likely evolved adaptively across millions of years for overcoming challenges associated with invading and entering demanding coral hosts.

Key words: adaptive host invasion, larval phenotypes, coral barnacle, barnacle phylogeny

CORAL BARNACLE COLLECTIONS

Live corals of Leptastrea purpurea and Psammocora profundacella bearing burrowing and adult females of Berndtia barnacles were collected by SCUBA diving on coral reefs in the vicinity of the Northeast (NE) coast of Taiwan. Adult specimens on approximately 3 cm² host colonies were transferred live to the lab and kept in 1-liter beakers containing aerated, filtered seawater with a 10L:14D cycle under LED lamps. The seawater was changed daily, and the tanks changed regularly. All tanks were checked daily for the release naupliar larvae (which precede the invasive cyprid larva). Once released, the nauplius larvae were concentrated using a pointed light source and then transferred using glass pipettes to sterilized aerated petri dishes. The larval material of the acorn coral barnacle Darwiniella angularis (Thoracica: Pyrgomatidae; inhabiting Cyphastrea chalcidicum) originate from Liu et al. (37).

Larval culture and imaging

Hatched nauplius larvae of Berndtia from the field-collected females were cultured in autoclaved seawater in petri dishes at ~26°C. As the larvae of Berndtia are lecithotrophic (non-feeding; Video 2) and no food was supplied during the culture. Seawater was changed at daily intervals until the nauplii metamorphosed to the cyprid stage (seen by the emergence of two compound eyes laterally to the nauplius eye; Fig 3A). To assess the larval swimming behaviors and gross structure in culture, both nauplii and cyprids were video recorded and photographed alive in either an Olympus SZX7 stereo microscope (SM) or a Zeiss AX10 light microscope (LM) fitted with differential interference contrast (DIC) optics.

All larvae investigated originated from mixed broods of various females. To gain higher morphological resolution of the larval structure, >50 nauplii (all instars) and >100 cyprids were fixed and prepared for LM and scanning electron microscopy (SEM) following the guidelines in (8, 29). Digital editing of images was done using Corel PHOTO-PAINT X8 and the photo plates were assembled in Corel DRAW X8. The videos were edited in ACDSee Photo Studio.

Larval experiments

To track how the Berndtia cypris larvae explore and invade their hosts experimentally, we released between six and 50 live cyprids from different broods in 10-litre transparent, rectangular aquarium tank containing host colonies. All tanks were aerated by air pumps and sterile glass pipettes. The filtered seawater was changed daily. The cyprids of all three species are non-feeding. The condition of the cyprids were checked every 10-15 minutes during daytime under a stereomicroscope (SM) after exposure to the hosts. The SM was fitted with a digital Lumix G8 camera. Live videos were recorded at 15-minute intervals for 4-30 minutes from about 06:30am-11:00pm but were on several occasions followed for 24 hours. This was repeated for 32 days for all three species. To document larval and juvenile metamorphosis, observations and videos were made every 30 minutes after larval invasion.

COLLECETION, REARING AND IMAGING OF OTHER BARNACLE LARVAE

We collected ascothoracid (Baccalaureus sp.; Video 1; Fig 6B), burrowing (Acrothoracica; Video 1) and acorn (Balanomorphasp; Video 1) cypris larvae by deploying light traps in Gongguan Harbor on the NE coast of Green Island, Taiwan. The plankton was sorted in Petri dishes under an Olympus SZX7 dissection microscope, and 10 ascothoracid, one burrowing and one acorn live cypris larvae were recorded in an Olympus IX70 light microscope. Five specimens of the ascothoracid cypris larvae were processed for SEM as described above.

            The material of Trypetesa lampas (Acrothoracica: Trypetesidae; Video 1), Peltogaster paguri (Rhizocephala: Peltogastridae; Video 1; Fig 1B) and Scalpellum scalpellum (Thoracica: Scalpellidae; Video 1; Fig 1C) were collected in the vicinity of the Kristineberg Marine Biological Station in Southern Sweden. Hermit crabs (Pagurus bernhardus) were removed, and their shells cracked open using a hammer. The shell pieces were examined under a dissection microscope for female T. lampas specimens (often identified by exhibiting a pink tint in vivo). The free hermit crabs were screened for infestation with P. paguri. The specimens of S. scalpellum were collected from thecate hydroids in the same area by dredging with an Agassiz-trawl. Adult specimens of the tree species were subsequently placed in small, aerated culture vessels. Upon release, the non-feeding nauplius larvae of several female/hermaphrodite broods of the three species were cultured in separate cultures with their conditions being assessed regularly. The cyprids were upon emergence photographed and live recorded under a LEICA DMRXA microscope fitted with an EVOLUTION© camera.

            Lastly, we live recorded live cyprids of two additional acorn coral barnacles, Galkinius altiapiculus and Trevathana savignium to put further emphasis on the larval phenotypes of acorn coral barnacles. The former (Video 12) was collected around Taiwan (Suao, Kending, Turtle Island, Green Island and Siao Liou Chiou Island). Small pieces of corals with adult female Galkinia barnacles were collected using a hammer and chisel at 0-20m. The females and their broods were processed as described above for Darwiniella angularis. The cyprids of Trevathana magaretae were reared from females attached to the coral Platygyra lamellina in the Gulf of Aqaba as described for Galkinia. The cyprids were reared as described for Darwiniellaangularis with the exception of using 20-24°C and at 13L:11D light cycle using LED-lamps. The nauplii are non-feeding and were not fed. Cypris larvae were live recorded using a Sony Power HAD and as for the Kristineberg-material.

COLLECTION, DNA EXTRACTION AND PCR OF Y-LARVAE

The Facetotecta is an enigmatic taxon for which no single-specimen voucher exists. To adequately root our trees outside the focal group of interest it was necessary to expand the marker coverage of the Facetotecta. We deployed plankton nets along the pier of Gong Guan Harbor, Green Island, Taiwan between 15:30-18:30 in September and October. Y-larvae were picked under an Olympus SZX7 stereo microscope with Pasteur pipettes and placed in small Petri dishes. +100 specimens of new, unnamed y-larva species (here referred to as Facetotecta sp. 9 to maintain consistency with the naming by (37)) were photographed alive and fixed in 95% ethanol. Additionally, we collected 163 specimens of Hansenocaris itoi Kolbasov and Høeg 2003 in the waters off the White Sea Marine Biological Station, Russia (66°34’N, 33°08′E). All larvae were captured by vertically dragging 72μm mesh net with a 40 cm mouth opening from 40-0 m. 17 naupliar specimens were fixed alive in 95% ethanol.

Eight specimens of Facetotecta sp. 9and 14 of Hansenocaris itoi were transferred individually to sterile Eppendorf tubes in 1.0uL ethanol drops. The DNA was then extracted by adding 40uL AE-Buffer and 4uL Protease K (QIAGEN™) and incubation at 56ºC for 1h and 75ºC for 12min. The exuviae of the specimens left at the tube bottom were subsequently mounted on glass slides in glycerine jelly and photographed in a Zeiss AX10 light microscope fitted with differential interference contrast (DIC) optics. The slides are stored at the Natural History Museum Denmark.

            We used the polymerase chain reaction (PCR) to amplify partial sequences of 12S ribosomal DNA (rDNA, 400bp), 16S rDNA (330bp), 18S rDNA (1100bp), 28S rDNA (770bp), cytochrome oxidase subunit I (COI, 650bp) and Histone-3 (H3, 330bp). Primers and PCR conditions are listed in Supplementary Table 2. We pipetted 0.4uL of each 10uM primer, 12.2uL ddH₂O, 4.0uL Fast-Run^TM Taq 5x Master Mix with Dye (Protech Technology Enterprise Co., Ltd., Taiwan) and 3.0uL DNA template into each Eppendorf tube. PCR was conducted in a DNA Engine Thermal Cycler (Bio-Rad, Richmond, California, USA), and the products were visualized by electrophoresis on 1.5% agarose gel in 1 × TAE buffer. DNA purification and Sanger sequencing were performed by Genomics BioSci & Tech Ltd. (New Taipei City, Taiwan). The sequences were assembled and edited in Geneious Prime 2020.1.1. We uploaded curated sequences to NCBI GenBank (Accession: XXXX-YYYY). These sequences represent the first single-specimen-vouchered nucleotide sequences of the Facetotecta.